[X] My Assistant

Hey, it’s me, Imageset Shitposter Man.
I’m here once again to ask for, well, maybe just a moment of your time.

I don’t know if you guys have noticed, but imagesets are becoming more and more prevalent.
I can't exactly give solid numbers as I haven't been keeping track that far back, but the amount of imagesets I see posted per day are drastically higher than when I first joined the site.

This is not necessarily a bad thing. Imagesets are good (sometimes). There's nowhere else on the site you can post an artist's miscellaneous art, much of which may even be deleted or otherwise not found anywhere else.

However, it is also true that imageset uploading is the primary method by which point farmers and toplist chasers operate.



This is the current top 10 positions in yearly toplist for uploading galleries.
5 of these make a bulk of their income from imageset-type galleries (saltcutlet, Pokom, ROBBOO, Godboli, Viktor0nk)

You can argue that uploading content should simply be rewarded in turn, and that it is not intentional point farming.
I believe this is true for the other half of the 10.

However, I disagree wholeheartedly in the case of the imageset uploaders.

I myself can admit that I've acted in the interest of gathering points in the past.
For the others, well, proving that one acts in the interest of point farming is difficult. You can make assumptions from how they upload or what they upload, but it is not concrete.
Nonetheless, there is one act that reveals that their interest is in points, not providing the gallery system with content.

To remove content for the sake of getting more views is solely the act of a point farmer.

Pokom


ROBBOO


Godboli


Viktor0nk


A bonus to show that it's not just restricted to the top end.

catchv


These are all cases where the uploaders have removed or expressed interest in removing content from their galleries in an attempt to de-fjord their galleries.
This does not include scenarios where an uploader purposefully excludes such content from being published in their galleries in the first place, as it cannot be easily proven that they did so. It is an easy assumption to make from the previous acts though.

Their intentions established, let's move on to why imagesets are their choice.

I covered a lot of it in my 2019 post.
I want to add on and re-iterate a bit though.

Namely, imagesets gain from being uploaded incomplete and later.

Some benefits of being incomplete are highlighted in my previous post. One more is depicted in the screenshots above, where the point farmers intentionally desire their galleries to be less complete for the sake of some tags.

However, what about being uploaded later?
Well, just give it a bit of thought.
The power of imagesets lie in their ability to abuse the frontpage. If another imageset is uploaded after, then the effectiveness of the first is lowered.
For the most part, this is true. Some exceptions apply, but after looking through galleries and testing, this is a general rule to follow.

Well, what is exactly wrong with these two traits?
Imagesets operate in the complete opposite as other gallery types.

Other galleries have protections in place to allow a higher quality gallery to replace a lower quality gallery.
Other galleries are expunged if they are equal or lower quality to one that was uploaded previously.
Imagesets are an exception to this.

https://ehwiki.org/wiki/Expunging#Different_Versions

All versions of the following may exist simultaneously: Imagesets of any theme, if uploaded by a different user. There is no such thing as a "more complete" collection in this regard.

In every other case, it is better to attempt to be the first to upload a high quality gallery.
For imagesets, it is the exact opposite; if possible, be the last to upload a gallery of any quality.

I can only think of 3 main ways to move forward from this.

1. Place restrictions on imagesets and the manner in which they are uploaded
This is not strictly viable as it places a lot of burden on moderating them. There is also not a solid way of comparing an imageset to another.

2. Place restrictions on the rewards of uploading imagesets
Likely an easier option, this can have the issue of harming smaller and less viewed imagesets even moreso.
It is also a much more "artifical" fix that may be responded with simply uploading more imagesets or content of other categories (which is honestly probably okay).

3. Do nothing
Accept the status quo and go with it.
Of course, I don't believe this is okay, which is why I made this post. Maybe others are fine with the way things are though.

Have a nice day.

I'm completely fine with cutting down the various point gains from Imagesets down to a fraction of what they are now. Hell, I wouldn't shed a single tear over them being set to 0 from here on out.

I'm fine with that option as well, but I don't know what the general user response would be.

I’m curious as to why this thread was restricted from viewing.
If any rules were broken, please let me know.
I’ll edit or repost with edits in the hopes of getting a discussion going.
I firmly believe that this is an important topic in the best interests of the site.

Wasn't my idea.

Neither mine - which leaves very few people as suspects.

Neither mine - which leaves very few people as suspects.

And as one of them, I can confirm it wasn't.

To be frank though only the administration can make these kinds of changes, so may as well only poke him.

Would be lovely be able to view who does what in a moderation sense, hopefully on the new forum..

Anyways, I very much would like to see the imgeset spam greatly reduced, and making them worth a lot less points would probably fix it to some degree, have anyone talk with Big T?

I already summoned him. It's his call on what approach (if any) is most feasible.

If people are bothered by image sets, they can just as easily turn off the category altogether, so I assume this really is just about the toplists. However, I don't think changing the toplist weight for image sets is feasible (or, at the very least, fair) at this point, seeing as it would probably have to be retroactive, so it would change the scoring for everything uploaded so far - nor would it necessarily fix anything.

Making changes to the rules of "completeness replacements" and suchlike probably isn't a good idea either, since we'd just end up with a bunch of replace wars on our hands.

Ultimately, I can see your problem, but I don't have a good solution for you. The gallery system wasn't really intended for continuously updated galleries like how image sets are being used here, but seeing as how those galleries are actually popular, disallowing them is not an option.

These are all cases where the uploaders have removed or expressed interest in removing content from their galleries in an attempt to de-fjord their galleries.

This does not qualify for a tag removal anyway. Since tags apply to a chain of galleries (i.e. the original gallery and all its updates), fjord content being present in *any* gallery in the chain qualifies it for said tag. Which means that even if the images with the content were "removed" with a gallery update, the tag would still be valid and must not be removed. Of course, the removal is an illegal action anyway, and would make the gallery subject to a rollback+lock if reported.

The gallery system wasn't really intended for continuously updated galleries like how image sets are being used here, but seeing as how those galleries are actually popular, disallowing them is not an option.

That is why I heavily favor reducing any gains that imagesets receive.
At the core of it all, imagesets operate outside the norms of the gallery system.
The "Misc" category captures content that is outside of the main intention of the gallery system and receives reduced gains.
The "Imageset" gallery type consists of content that operates out of the norm of the galleries.
Both are allowed, but only one is discouraged (ignoring the intersection of content such as 3D imagesets).

The examples of blatantly abusive actions done by top imageset uploaders I gave are just to establish the premise for my argument that point farming through imagesets is their priority, and thus, reducing their point gain is a viable method.

As one of (if not the) main recipients of the rewards of imageset uploading, I at least believe that it is fair to go with that method.
If it must be retroactive, I think that's alright as well.

I do not think that there is a good solution within reach either.

However, action should be taken now before more of the gallery system is bogged down by continuously-updated imagesets.
If imagesets can not have their content regulated, then the next best option is to have incentive for users to put effort into their galleries.
That is also not 100% possible, as determining "completeness" of an imageset-type gallery is not feasible and/or requires tremendous effort for those that would judge it.
A step down from that is to penalize such gallery types as a whole, so that those who would be interested in abusing lower quality uploads turn to gallery types that can actually be regulated. In turn, it would leave uploaders who are actually in it to provide content.

The popularity of such content would not be affected on the end of the viewers. In fact, it may be the opposite as higher quality galleries arise because those involved are not in it for toplist farming.


(In my opinion, an ideal solution would be a booru-like system specifically for loose content which aggregates such uploads, and rewards uploaders for images contributed.)
(After all, such content is mainly handled by boorus.)
(However, to add that to E-Hentai would require tremendous development time and be somewhat redundant)

This post has been edited by saltcutlet: Jul 30 2020, 16:15

I think I asked before but how about non-bumping gallery updates?

That might help, but it would still introduce abuse.
Maybe more so, as it would give incentive to uploaders to be the last to bump their galleries, knowing that others would not have a chance of doing so afterwards.

Another idea I had was to hide imagesets by default on the front page and/or searches.
If the user wants to see them, they’d have it enable it.
Alternatively, hide imagesets on the first X pages.

This post has been edited by saltcutlet: Aug 12 2020, 15:56

So, is there any chance this could get unrestricted?

I'd rather an open dialogue and more ideas from more people.

By default I'm hiding those galleries, given I mostly don't care. Lately I've starting hoarding Pixiv artist, and I myself uploaded some galleries.
However I don't use the galleries here, because I often (not always) find them incomplete which goes toward the problem first mentioned here: the chase of the rewards.

Here my ideal rules:

1. Forbid any upload incomplete: the upload should have all the pictures from the sources (Pixiv/Twitter/DeviantArt/etc.). And thus allow the incomplete tag for imageset.
2. Upload chronologically: I find it annoying, to say the least, that they upload pictures randomly.
3. Decrease rewards: those galleries are getting a lot of update, even only because some artists produce a lot. This artificially give them more view.
4. People abusing the system should be punished (ban of imageset upload): examples in the OP of saltcutlet, having incomplete galleries for the bump shouldn't be allowed.
5. The rule about 1 uploader and 1 gallery with different sources is great.

I've read 10b answer so my ideal rules are not fully possible, also this will put more load into the Vigilante. But I don't see how to handle this in another matter. Would it be possible to have a way to extract all the poster of imageset filteed on galleries having [Pixiv] in the title? I'm not sure, but I don't think we have that much uploaders. When I looks quickly I seen like ~5-10 people massively uploading.
If we take actions so that they'll respect new rules, this should at least be a bit better IMO.

I must also say that my view on e-hentai is about having a place with organised picture. If I want to browse pictures "randomly" there are way better way (*booru). That's why I dislike most of those galleries and barely use them, it's a mess. Though lately some changed their way (like Pokom) and finally upload them in an organised way. My problem is that we're going into having the whole Pixiv & co. into e-hentai, I'm not sure if this is a good idea.

I liked the idea of Maximum_Joe: non bumping update. I don't understand the abuse you're talking about saltcutlet though? If they upload the gallery one time, whatever the updates they'll do it'll be invisible.

I’ll just repost what I said over the GSC bc I’m too lazy to do paraphrasing:

reduction of imagesets gains

Think I have said this before over the IRC, but despite the flood of nonsenses, there are still legitimate imageset galleries out there. Some Few users still upload “meaningful” content (like Patreon-exclusive images or something that), and punishing these innocent ones as well is just a bit overkill imho.

If everyone gained near-nothing, then there is no competition to encourage cheating the system.

What worries me the most with this nuke-em-all solution is whether the GP farmers will revolt by just simply spamming even more worthless boorus around here.

That being said, I do support the idea of having non-bumping updates.

This post has been edited by PrincessKaguya: Aug 30 2020, 12:40

OT: You have also included an uploader that intentionally uploads low quality doujin, and then updates to high quality a few days later; every time, for the sole purpose of milking points.

What worries me the most with this nuke-em-all solution is whether the GP farmers will revolt by just simply spamming even more worthless boorus around here.

I agree that's why I suggested more rules. What worry me is more the incomplete galleries, if they do spam galleries complete and well organised, that's not much of a problem. At least those galleries will be good, even if they're worthless for most of people.

I think we also have two problems there, which different way of resolving them
* GP farming
* "Bad quality" galleries

I've read 10b answer so my ideal rules are not fully possible, also this will put more load into the Vigilante. But I don't see how to handle this in another matter. Would it be possible to have a way to extract all the poster of imageset filteed on galleries having [Pixiv] in the title?

I'm not 100% sure if this is what you're suggesting, but one of my ideas was for the gallery system itself to collect miscellaneous images into one gallery for each artist.
It was mentioned in more detail in my GSC thread.
There are a lot of logistics problems even with that, but IMO this is an ideal way to handle such content.

I must also say that my view on e-hentai is about having a place with organised picture. If I want to browse pictures "randomly" there are way better way (*booru). That's why I dislike most of those galleries and barely use them, it's a mess.

That's basically the crux of the issue. The E-Hentai gallery system and its rules were created and updated to reflect works such as doujinshi and sold CGs.
Boorus are meant to handle such miscellaneous content and a lot of rules have to be readjusted just to pander to loose works that go into imagesets, resulting in the ability to abuse them.

If so many accommodations have to be made just for such content, there's no reason for them to earn as much as other types.

I liked the idea of Maximum_Joe: non bumping update. I don't understand the abuse you're talking about saltcutlet though? If they upload the gallery one time, whatever the updates they'll do it'll be invisible.

Here's an example:

If we introduced a 30-day no-bumping limit...
And we have three uploaders for one artist X...

Uploader A uploads his gallery on 2021-01-01. It cannot be bumped until 2021-01-31.
Uploader B uploads his gallery on 2021-01-03. It cannot be bumped until 2021-02-02.
Uploader C uploads his gallery on 2021-01-10. It cannot be bumped until 2021-02-09.

In this first month...

A's gallery is the first imageset for the artist viewers can find for 2 days, until B's upload.
B's gallery is then the first for 7 days, until C's upload.
C's remains at the top as no other gallery was published, and remains so for the rest of month until A's cooldown is over.

This cycle repeats every 30 days.

OT: You have also included an uploader that intentionally uploads low quality doujin, and then updates to high quality a few days later; every time, for the sole purpose of milking points.

I'm aware. However, this can only be done a handful of times at best, unlike imageset-type content.

This post has been edited by saltcutlet: Aug 30 2020, 19:24